Dynamic Allocation of FlexE Discrete Resource Based on Regional Traffic Prediction

Flexible Ethernet (FlexE) can ensure hard isolation of network slicing and match flexible rate. We consider the process of network slicing transmitting from the radio access network to the transport network. However, resources are allocated discretely in FlexE and they have different granularities....

Full description

Saved in:
Bibliographic Details
Published in2022 3rd Information Communication Technologies Conference (ICTC) pp. 86 - 90
Main Authors Hu, Yuqi, Ma, Deyu, Sui, Jingqi, Yao, Lei, Bai, Haonan, Zhang, Yong
Format Conference Proceeding
LanguageEnglish
Published IEEE 06.05.2022
Subjects
Dynamic scheduling
FlexE
Graph neural networks
hard isolation
Heuristic algorithms
Network slicing
Prediction algorithms
Quality of service
resource allocation
Simulation
traffic prediction
Online AccessGet full text

Cover

More Information
Summary:Flexible Ethernet (FlexE) can ensure hard isolation of network slicing and match flexible rate. We consider the process of network slicing transmitting from the radio access network to the transport network. However, resources are allocated discretely in FlexE and they have different granularities. Therefore, we propose a dynamic deployment method of discrete resources based on regional traffic prediction, which comprehensively considers the delay and bandwidth requirements of different slicing services, while improving the overall resource utilization and reducing the traffic loss rate. Simulation results show that the proposed algorithm can improve resource utilization and reduce traffic drop rate while meeting the service delay requirements.
DOI:10.1109/ICTC55111.2022.9778402